SNP Tracker

Nothing to see here until you type or choose a SNP to view.

Westward Migration

notes ↓
Like North America, the population of Western Europe has been shaped by migration from the east -- but multiple times and thousands of years earlier. This chart shows longitude vs time to help visualize these migrations. The colors and thick solid/dashed lines are the same as the map, and the thin horizontal dotted lines show south-to-north lines at notable longitudes.

Note that haplogroup I precedes nearly all of the others; it can be found in the west during the Paleolithic, staying south of the glaciers of the last Ice Age.

Population Over Time

notes ↓
This chart shows the estimated number of men who have a given SNP, soon after that SNP appears. The estimate is based on a very simple growth and split model: each population grows at the known exponential rate for its era, and then when a clade splits, the population is divided in proportion to the future population with that SNP. Thus the drops in this chart do not indicate plagues or wars, but simply clades splitting into subclades. By construction, adding up all of the clades would reproduce the known exponential growth of the entire population. Modern-day terminal SNPs have small populations as they must, due to recent creation and limited sampling.

It is very intriguing to note how often an unbranching population stays near the Dunbar Number of 150.

Speed of Migration

notes ↓
Since we have dates and locations for the SNP dots on the map, we can calculate the average speed of migrations. All of the usual caveats about accuracy apply, especially for anything in the last 2000 years, but the older speeds should be reasonably accurate. Note the log time scale so that you can see the whole path.

Most paths are boring: Haplogroup Q for example trudges from Africa and back to Sweden or Pakistan at a leisurely average of about 0.2 kilometers per year. Haplogroup I moves west across Europe, south of the glaciers, at the same pace. Of course no one should think of constant migration: groups probably settled down for centuries and only moved on when hunting, grazing, competition, or other forces made them move.

What I had hoped to see shows up nicely, namely the speed of nomadic cultures with horses and carts. Look at any R1b subclade and note the speed between SNPs R-L23 to R-P312, averaging about 2 km/year up the Danube valley -- ten times faster than the pace of Paleolithic tribes on foot.

Nothing to see here until you type or choose a SNP to view.

notes ↓
For SNPs in a path:
  • If you've entered multiple SNPs, only the first is shown here.
  • The number of descendants counts men who have done Y DNA testing with FTDNA.
  • The anchor symbol indicates a SNP with location established by the archeaology literature, not dependent on user-reported ancestry. SNPs between anchor points are interpolated assuming a constant rate of travel.
  • The six most prevalent countries are shown in each row for SNPs after the last anchored SNP. A weighted average of these countries determines the SNP dot location on the map.
  • The legend below the main table identifies country icons, and its bar graph shows contribution to locations, weighted by time (most recent with greatest weight).
  • All dates are shown to only two significant figures; the statistical uncertainty of SNP counting does not merit any greater precision.
For SNPs associated with a surname:
  • These SNPs are unrelated -- they are not a path to a single endpoint.
  • The SNPs are listed oldest-to-youngest only for convenience.
  • SNP country data is not broken out by surname, and so is shown only when all of those with the SNP have the same surname.

How to use this tool

Use the (/) toggle button to choose Y DNA () or mitochondrial (mt) DNA ().

Simply type your haplogroup (ideally your terminal SNP) into the input box and click "Go" or hit return. If this SNP is known to the FTDNA Y or mt Tree, the map will show the path from "Y Adam" or "Mitochondrial Eve" to the SNP. You need type only the lower-level code, for example "M222" instead of the full "R-M222". Y SNP names are as in FTDNA, not the older nomenclature, thus "I-L160" and not "I2a1a1" (although many short 'classic' labels will work, like R1a or G2). If what shows at the top-left of the map is different from what you typed, that's because the tool automatically searches all SNP synonyms and displays the current FTDNA nomenclature. If your haplogroup is missing, then it's not in the FTDNA Y or mt Tree. You may also select examples from the drop-down menu or tap the "shuffle" button ( ) for a random SNP. You can compare paths by entering multiple haplogroups, separated by commas.

Hold down the option/alt key and click ( ) for a random surname (which applies only to Y DNA).

Map Options

Once you have a path showing you can open the options panel (click ).

"Zoom to Europe" toggles between views of Eurasia/Africa and Europe. The camera button ( ) sends a JPG file to your Downloads folder (at least on Safari, Chrome, and Firefox; I have no way to test IE or Edge). The "Smooth Path" toggle optionally invokes an algorithm that removes much of the scatter of self-reported locations while trying to be consistent about traversal time.

"Show " will drop down a simple animation slider control. Click the play arrow to start the animation of a walking man who will trace your paternal or maternal ancestry. You can pause the animation () and then drag the slider to place the walker anywhere on your path. Relevant events and cultures will appear and disappear as the walker passes by.

"Show Topography" toggles between a minimal coastline background and an topographic map. The topographic map was generously created Tom Patterson; he and his and colleagues at Natural Earth ( and produce beautiful maps that show the earth without human labels or influence.

"Show Descendants" will display all of the descendant SNPs of the last SNP in your path. This has no effect if your path ends in a terminal SNP, but it gives dramatic results with major ancestral SNPs such as F-M89 (ancient Mesopotamia), I-M170 (associated with Western Hunter-Gatherer), R-M417 (Eastern Hunter-Gatherer), R-L23 (Yamnaya), and I-M253 (early Scandinavian). The descendant SNP dots have the usual color code, so for example with F-M89 you can see the gradient of migration northwest over the millenia. The strong selection bias of FTDNA testers is also evident in the Finland-to-Britain dominance of locations.

What to look for

Notice how dramatically different our paths have been from Africa: Try the examples with multiple paths to England, the Isle of Man, and Finland. Your next-door neighbor's ancestors might have taken a prehistoric path thousands of miles and years different from yours.

Find long loops and round trips. Compare R1a to R1b: They're together until they get out to Mongolia and Lake Baikal, then split with R1a taking the northern route west to Europe, while R1b retraces its ancient path and weaves around the Caspian and Black Seas, finally following the Danube valley into Europe.

Look for the evidence for "starbursts" of dramatic expansion. The I1 Viking, I1 Norman, N1c Finnish, R1a Polish, and all R1b paths show a cluster of slate-blue Neolithic SNPs in close proximity: that many SNPs in the same region means that people stayed put with a fast-growing population.

The Mixed Blessing of Averages

At some point, averaging data to estimate locations simply gives silly results. A Spanish mother and German father doesn't make me French (geographically halfway). And yet there is power in numbers and averages: if a certain SNP has 10 descendants who say that they're Irish and 2 who say something else, and you have that SNP, the probability of you having Irish ancestry is significant.

For many people the Neolithic is a sweet spot: severe bottlenecks followed by strong population expansions resulted in a few very big SNPs like R-U106, R-P312, and I-M253. Descendants of these will see a fairly tight cluster of blue (Neolithic) dots in one region. High numbers lead to reasonable averages and often smooth downstream paths to Finland or Scotland.

Many Middle Eastern SNPs are a middle ground: the history of J-FGC2 is scattered from Mesopotamia to Anatolia in the Paleolithic and Neolithic, then settles down to Saudi Arabia (probably with testing bias). SNPs with lower testing levels are the most scattered: paths from T, G, and J2 tend to be poorly determined.

Also beware of the scatter associated with large countries: Russia and China get a single central location despite their huge geographic size. A single Russian descendant can pull a SNP point many hundreds of miles across Europe, even though that Russian ancestor might have lived just over the Estonian border.

Data Source for Dates and Sequence

This tool depends on the generosity of Family Tree DNA (FTDNA) in making their Y SNP Tree data accessible. Note that there is no personal data involved (no kit numbers), only the phylogeny of the tree and self-reported ancestor locations. The initial versions of this tool used my own SNP dates, but since the Y Tree data feed does not include private variants, many SNPs got dates that were much too recent. The tool now uses dates from YFull: about a quarter of the SNPs in the Y Tree have corresponding dated SNPs in YFull which suffice for interpolation to cover the entire tree.

SNP formation dates in YFull are estimated by the method of Adamov et al . Since SNPs are random and independent, such dates have intrinsic error which is shown in the table as 95% confidence limits (see also graph at right and Adamov Table 5). Note that recent SNPs have very large uncertainties -- accurate dating in this range must rely on STR statistics or paper genealogy.

Self-Reported and Historic Locations

Map locations are intended to show where a given SNP mutation occurred, not where a haplogroup may be most prevalent today. Locations are estimated with a combination of (1) averages of FTDNA testers' reported ancestry for over 20,000 Y or 5400 mt SNPs, and (2) about 150 specific SNPs that I have anchored based on academic scholarship. These sources are complementary: paleolithic dates are very poorly estimated by modern testers' history but a small number of academically-located paleolithic SNPs suffice for the most populous branches. On the other hand, averages of modern reported ancestry are often quite good for medieval to modern localization.

Wikipedia and its references are a major source for academic SNP locations. The Indo-European project and Eupedia are also excellent sources with maps, references, and detailed discussion. Most migration maps (for example , ) are too general to be useful in locating specific SNP origins in time and location, though for mtDNA there is not much more (see below).

Carlos Quiles recently shared his hand-curated list of over 8700 ancient human remains with locations, dates, and Y and/or mtDNA haplogroups. But even with this trove of data and five different approaches, automated methods relying solely on ancient sites are no match for human interpretation. See for details.

I hand-curated a few anchored points where the given locations would have required a very unusual rate of migration. It is also worth noting that a number of published ancestral tracks are probably incorrect because they do not fully account for the time-sequence of SNPs. Every path on this map is drawn by traversal of the Y or mt SNP tree and thus always in correct past-to-present order.

Personalized Terminal SNPs

As more men do BigY DNA testing the Y SNP tree has grown to the point where many terminal SNPs now overlap the timescale of paper genealogy; this allows those testers to pinpoint the precise origin of that SNP in time and geography. SNP Tracker identifies these SNPs and fixes those that have a single country location. In addition, I will add more specific data if asked. Use the button below to email me your SNP name, surname, SNP formation date, and origin latitude and longitude in decimal degrees, not degrees/minutes/seconds; west longitude is negative. Note that only European or Asian locations will show on the map, so even if your terminal SNP is actually in North America, give me the last known European location. Only BigY terminal SNPs please: I will not modify any SNP with more than one surname attached, or any earlier than about 1800 CE. I generally do data updates every 4-6 weeks.

Ask for Custom SNP Data

Ancient Global Consistency and Recent Smoothing

Since November 2019 ancient SNP paths are computed as a whole, taking account of their interdependencies, and recent paths may be optionally anchored to very recent terminal SNPs that cite a single ancestral country. The latter change recognizes the increase in BigY data which means that many customers' terminal SNPs now extend (in time and geography) into the range of paper genealogy. See the Research Report for details.

Mitochondrial DNA
Ancestral paths for mtDNA are much more difficult to estimate than for Y DNA: mtDNA mutations are fewer and rarer (one per ~10,000 years, vs one per ~100 years for Y SNPs or Y STRs). mtDNA is much more diverse since maternal lines have not had the purifying effect of major bottlenecks (near-extinctions) like Y DNA, thus mtDNA does not have easily traced "migration superhighways" like Y DNA (i.e. R1b-M269, R1a-M198, I1-M253). As a result, many mtDNA lineages tend to be ancient and found across all of Eurasia. The literature offers less information about founders and splitting, so the mtDNA paths here are largely based on continent-scale general migration maps such as and to a lesser extent .
Why such poor coverage for Africa, India, China, Oceania, and Native Americans?
The tools depend on location and phylogenetic data from FTDNA testers, which is currently too limited for these locations. The majority of FTDNA testers are Americans of European ancestry.
Caveats, questions, comments, requests

This tool is entirely free and it comes with no guarantees. I have tried to make the maps as accurate as possible, but the more recent points on any path will depend on self-reported countries of ancestry. They will change as new data arrive. If and when FTDNA adds dates to their Y Tree data, I will adopt those since they're likely to be more accurate than my method of interpolation from YFull.

If you think that I got something wrong, particularly in paleolithic locations, please let me know; if you include research and references to back up your assertion, then I will modify the paths. Remember that I cannot control recent (post-Roman) locations because they're based on FTDNA user data.

Please tell me about your ideas for improvement and stories where this tool has been of particular interest to you.

Rob Spencer    
Contact Me

Choose "Show " under the Map Options menu ( ), and a walking man icon will traverse the path of your paternal or maternal ancestors. Selected major events and cultures appear as it passes by; thanks to Wikipedia this page has notes on each. The haplogroup buttons below show example SNPs which first appeared at the same time and place as each event or culture -- but note that haplogroups are not exclusive: all cultures were mixtures.
Loading Data